A flocking based method for brain tractography.

Area of Science:

• Neuroimaging
• Computational Neuroscience
• Biophysics

Background:

• Accurate estimation of axonal fiber pathways is crucial for understanding brain connectivity.
• Existing diffusion MRI tractography methods face challenges with noise and complex fiber crossings.

Purpose of the Study:

• To introduce a novel method for estimating axonal fiber pathways using Multiple Intra-Voxel Diffusion Orientations (MIVDO).
• To enhance the robustness and accuracy of brain connectivity mapping through improved tractography.

Main Methods:

• Utilized stochastic particle walks incorporating mass, gravitational, and inertial forces to model fiber trajectories.
• Implemented a gravitational interaction, akin to flocking behavior, for collective particle movement.
• Applied principal component analysis and spectral clustering to filter outlier pathways.

Main Results:

• Generated smooth, filtered, and compact axonal trajectory bundles.
• Demonstrated improved robustness in axon fiber estimation through collective particle information.
• Successfully filtered outlier pathways, reducing false connections.

Conclusions:

• The proposed method offers a robust approach to axonal fiber tractography using MIVDO data.
• The flocking behavior and filtering techniques enhance the reliability of brain connectivity mapping.
• The method shows promise for both phantom data and in vivo human brain imaging.